Professional Documents
Culture Documents
Abstract
The end-windings of large generators are exposed to some of the largest vibrations among all
machines. The stability of these end windings has a major impact on the reliability of generators.
A resonant vibration condition can lead to movement and forces that can compromise the stator
bar insulation and possibly cause a complete failure of the generator. Currently, there is not a set
of established industry standards or acceptance criterion for generator end winding vibration.
There are many differing points of view within the power industry concerning the sources of
vibration, methods of analysis, and solutions. This applied research paper investigates generator
end winding vibration in order to establish a foundation to further the discussion among industry
and academia in seeking to form a set of standards. This project relates directly to the problem of
synchronous generator end-winding vibration at the J.H. Campbell power generating facility
located in West Olive, Michigan. There were significant concerns that the end-windings for the
A and B generators at the Unit-1 power plant were experiencing excessive levels of end-winding
vibration and were at risk of a major failure. The objective of this capstone project was to
perform an overall assessment including a determination of the cause of vibration, vibration
testing, data acquisition, and analysis of the end-windings for each machine. It was concluded
that the resonant vibration at a frequency equal to the fundamental electrical frequency as well as
harmonic components of the fundamental frequency is the root cause of the end-winding
vibration for each synchronous generator. Based on the analysis of test data a long term solution
was established and executed successfully to reduce the vibration problem. This is an industryuniversity collaboration that is very important to enhance engineering education and research.
Introduction
The vibration of generator end windings has been a topic of concern since the beginning of
power generation. Current flowing in the rotor and stator give rise to magnetic fields. The
resulting forces lead to vibration within the stator core, but more seriously at the stator end
windings and their support structures. The end winding structures have a high susceptibility to
vibration damage on account of the complexity of their structure, the number of materials used in
their construction, and the difficulty in supporting them properly [1]. Over the course of
evolution of the synchronous generator there have been many changes that had an impact on the
vibration of the system. The science of precise vibration monitoring and analysis is also a
developing technology which until recently was not able to produce accurate and accepted data.
The result has been a lack of consensus within the industry which has led to no accepted methods
or standards on the subject. This paper will discuss the sources of vibration in the generator and
the methods used to analyse them. The focal point of the discussion will be on stator endwindings.
Proceedings of the 2014 ASEE North Central Section Conference
Copyright 2014, American Society Engineering Education
(3)
where, = permeability of the core, N = number of turns, i = current in the windings, A = cross
sectional area of the core, and lc = length of the conductor. The power loss due to heat in a
conductor is expressed by (4) where R is conductor resistance. The ultimate result is a substantial
(4)
increase in heat as power requirements are raised. The increased heat levels within the endwinding conductors created a need for insulation materials to change over time. The type of
insulation has an impact beyond just thermal resistance. The damping characteristics from one
material to another can vary drastically which has an impact on vibration resistance. The early
insulator materials used for generator stators were organic asphalt mica. In general, organic
materials soften at a much lower temperature and have a much lower mechanical strength than
copper or steel. Thus, the life of a stator winding is limited most often by the electrical
insulation rather than by the conductors or the steel core [4]. The result was positive for vibration
dampening so the vibration problems did not arise as often. The change from organic materials to
thermo-set resins which can withstand much higher temperatures also had an impact on the
natural damping characteristics. The thermo-set resins are very rigid materials and also
susceptible to vibration problems. The overall result was an increase in failures in insulation
from vibration
impart a broad flat spectrum of frequencies over the range of interest. When this occurs the
component will begin to vibrate at its own natural frequency. This natural frequency response is
detected by a small sensor temporarily attached to the component. The output of the sensor is
fed to a structural dynamics analyser where the time response signals are filtered, averaged, and
converted to frequency response signals using FFT. Each component may be tested in multiple
locations along radial, tangential, or axial directions. However, this method has following
limitations such as: (i) during impact testing the winding will be at ambient temperature, and the
properties of all the non-metallic materials in the winding will be room temperature properties.
Since these properties are temperature dependant, room temperature behaviour is only an
approximation. (ii) the windings has not expanded and undergone any stress and shape changes
due to operation, and (iii) the forces applied during impact testing are very small compared to
those that will be experienced during operation. Therefore, the effects of any non-linearity in the
response of the winding to a forcing function will tend to be lost. Despite these limitations,
impact testing is the simplest way to evaluate the mechanical behaviour of the stator endwinding, and is capable of revealing a great deal of important information regarding the
performance of the winding. Most of the concerns listed above as limitations can be
accommodated by establishing a margin between the response measured during impact testing
and the known forcing frequencies during operation.
It is worth noting that a liquid cooled generator has a significant mass of water in the winding
during operation, and that this may have some effect on the winding response. If a suitable
source of clean de-ionized water is available, then this mass can be introduced into the winding
prior to impact testing, eliminating this source of uncertainty. Impact testing should be
performed on the fully assembled, cured winding structure. The global behaviour of the winding,
including natural frequencies, and mode shapes should be determined. Individual components
such as series loops, phase connections and phase rings should be tested. If natural frequencies
close to 60Hz and 120Hz in 60Hz machines are found, then mode shapes should be determined
to aid in design of suitable blocking systems. The basic goal of the impact test is to provide the
amplitude of vibration through a band of frequencies. The output of the impact test is to create a
graphical representation for the frequency signature.
The modal analysis test
Modal analysis is the term used to describe the modelling of patterns of vibration or mode
shapes for a given structure. If a structure has a natural frequency of vibration at or near its
driving frequency of vibration, this frequency can potentially be excited during operation,
resulting in wear and/or cyclic fatigue of the components. In order to reduce the effects of
vibration, structural analysis of equipment can be used to predict the mode shapes and make
structural modifications as required. Modal testing is a technique by which actual structural
responses relative to a measured excitation are recorded and used to create a modal model of the
structural response. The measured excitation may take the form of a single impact, a harmonic
shaker, or a reference signal representing the operational excitation. The general tendency of a
typical 2-pole generator end winding structure is to be deformed into an oval shape by the polar
forces of the generator rotor. The resulting vibration of the end-windings thus becomes that of a
deforming and rotating oval pattern. Because of this global structure vibration tendency, the
stator coils, and in particular the end of the stator coils in the area of their end connections, have
a resulting vibration which has its biggest component in the radial direction, that is moving
towards and away from the centre of the generator. It should be pointed out that the oval mode of
D.
an end winding is not the only mode that can be excited by forces within the generator. There are
other modes such as, for example, cantilever modes (the whole cone bouncing up and down like
a diving board) or breathing modes (the winding ring expanding and shrinking diametrically)
that could become resonant if forces act on the winding in the critical directions and at the
critical frequencies. The oval mode shape however is the most critical for global vibration
analysis of the stator because they get naturally driven by the rotor forces if the resonant
frequencies are close to the rotor forcing frequencies.
.
Conclusions
The performance and reliability of the stator end-winding system in a generator depends upon
the physical design of the end-winding, its support system, and on the forces that the system is
subjected to. There is no single method that can be used to diagnose a generator as having
excessive vibration levels. The use of expert visual inspection, impact testing, and modal
analysis should all be used together for a proper diagnosis. The use of online vibration sensors is
recommended if available but not required. There are many factors beyond the scope of this
paper that can be applied to the analysis of end-winding vibration. The overall intent of this
capstone project to provide a foundation for promoting the discussion has been presented with
the hope that an accepted set of standards will eventually follow.
Bibliography
[1] Shally, D., Farrell, M., Sullivan, K., "Generator end winding vibration monitoring," Universities Power
Engineering Conference, 2008. UPEC 2008. 43rd International , vol., no., pp.1-5,14Sept.2008doi:10.1109/UPEC.2008.4651488
[2] Klempner, Geoff; Kerszenbaum, Isidor (2004). Operation and Maintenance of Large Turbo-Generators. John
Wiley & Sons.
[3] Maughan, C.V.; Emeritus, P.E.; , "Vibration detection instrumentation for turbine-generator stator
endwindings," Electrical Insulation Conference, 2009. EIC 2009. IEEE , vol., no., pp.173-177, May 31 2009June 3 2009
doi:10.1109/EIC.2009.5166339
[4] Stone, Greg C.; Boulter, Edward A.; Culbert, Ian; Dhirani, Hussein (2004). Electrical Insulation for Rotating
Machines - Design, Evaluation, Aging, Testing and Repair. Wiley-IEEE Press.
[5] Humer, M.; Vogel, R.; Kulig, S.; , "Monitoring of generator end winding vibrations," Electrical Machines,
2008. ICEM 2008. 18th International Conference on , vol., no., pp.1-5, 6-9 Sept. 2008
Proceedings of the 2014 ASEE North Central Section Conference
Copyright 2014, American Society Engineering Education
[6] "IEEE Guide for Diagnostic Field Testing of Electric Power Apparatus - Electrical Machinery," IEEE Std 62.22004 , vol., no.,pp.0_1-100,2005doi: 10.1109/IEEESTD.2005.96280